Proc. Natl. Acad. Sci. USA Vol. 83, pp. 513-516, January 1986 Neurobiology Estrogen formation and binding in the cerebral cortex of the developing rhesus monkey (sex determination/hormone receptors/brain differentiation) N. J. MACLUSKY*tt, F. NAFTOLIN*, AND P. S. GOLDMAN-RAKICt Departments of *Obstetrics and Gynecology and tNeuroanatomy, Yale University School of Medicine, New Haven, CT 06510-8063 Communicated by Clement L. Markert, September 12, 1985

ABSTRACT These studies were undertaken to determine present studies demonstrate that estrogen receptors and the whether estrogen receptors and the microsomal enzyme system microsomal enzyme system called the aromatase complex, called the aromatase complex, which is responsible for con- which is responsible for conversion of androgen to estrogen, version of androgen to estrogen, are present in the brain of the are present in the brain ofthe rhesus monkey during perinatal rhesus monkey during perinatal life. Four monkeys (three development. The regional distribution of aromatase com- females-one fetus removed on day 153 of gestation and two plex activity in the cerebral cortex is consistent with a infants, 5 and 6 days postnatal-and 1 male, 2 days postnatal) putative action oflocally formed estrogen in the development were studied. Cytosol estrogen receptors were detected in all of areas of the association cortex concerned with cognitive brain regions examined. The apparent equilibrium dissocia- processes. tions constants for reaction of these sites with [3H]moxestrol were similar to those for uterine and pituitary cytosol estrogen receptors (0.3-1.1 nM). Within the brain, highest levels of MATERIALS AND METHODS binding were observed in the hypothalamus-preoptic area, Four monkeys (three females-one fetus removed on day 153 with fairly even, lower concentrations throughout the cortical of gestation and two infants, 5 and 6 days postnatal-and 1 structures. Aromatase complex activity was detected in the male, 2 days postnatal) were used in these studies. The fetus majority of the tissue specimens. The highest levels of estrogen was removed by caesarian section: the pregnant monkey was formation were observed in the hypothalamus. However, the sedated with ketamine (3 mg/kg of body weight) and anes- amygdala, the hippocampus, and several of the cortex samples thetized by intratrachial administration of a halothane/oxy- also contained measurable aromatase complex activity. Among gen mixture. The brain was removed from the cranium the cortical samples, the highest levels of aromatase complex immediately after delivery of the fetus from the uterus. The activity were found in regions of the association cortex (the postnatal animals were anesthetized with sodium pento- dorsolateral-prefrontal, orbital-prefrontal, anterior cingu- barbital (35 mg/kg ip.). Cerebral circulation and respiration late, and parietal cortices). The lowest levels of aromatase were maintained while a craniectomy was performed. The activity were found in the somatosensory and motor cortices of brain was removed within 12 min after sedation. the postnatal animals. These results suggest that locally-formed In all cases the brain was placed in a bath of ice-cold saline estrogen may be involved in the effects ofcirculating androgens for an additional 2-5 min after removal and then dissected on on the developing primate neocortex. a bed of crushed ice. Hypothalamus-preoptic area, amyg- dala, hippocampus, dorsolateral-prefrontal cortex, orbital- In many mammals, sexual differentiation of the brain results prefrontal cortex, parietal cortex, motor cortex, somatosen- primarily from sex differences in gonadal steroid hormone sory cortex, and visual cortex were dissected as described secretion during early life (1). Virtually all of the information (14). The anterior cingulate cortex was taken as the anterior about effects of gonadal steroids on the developing nervous half of the cingulate gyrus between the level of the genu and system pertains to differentiation of diencephalic structures the splenium. that underlie gender differences in gonadotrophin release and Cytosol estrogen receptors were assayed by using the reproductive behavior. Evidence for direct effects of the synthetic estrogen [3H]moxestrol (11P-methoxy-17a- gonadal hormones on the maturation ofother brain structures ethynyl-estradiol) to label the receptor sites and Sephadex is much more limited (2) and has never been reported for LH-20 gel filtration to separate bound and free steroid as primates. There is, however, abundant behavioral evidence described (15). Tissues were weighed and homogenized in 10 for sex differences in cerebral function in both humans and mM Tris buffer containing 1.5 mM Na2EDTA, 1 mM dithio- nonhuman primates. In the rhesus monkey (Macaca threitol, and 10% glycerol adjusted to pH 7.4 with hydro- mulatta), sex differences have been reported in play behavior chloric acid (TEGD buffer). The homogenates were centri- and in the effects of neocortical lesions during the first 3 fuged at 1°C for 1 hr at 105,000 x g. The supernatant cytosol months of postnatal life (3-6). In humans, sex differences fractions were incubated with 0.1-2 nM [11f3-methoxy- have been reported in cognitive abilities, cerebral lateraliza- 3H]moxestrol (New England Nuclear; specific activity, 79 tion, play behavior, and incidence of a variety of childhood Ci/mmol; 1 Ci = 37 GBq) overnight at 0-4°C. Parallel learning disorders (7-11). Circulating androgen levels during incubations, used in correcting for nonspecific binding, early life have been implicated as a possible causative factor contained a 100-fold molar excess of unlabeled moxestrol in for at least some of these sexual dimorphisms (11, 12). addition to the labeled steroid. Macromolecular bound ra- In many subprimate mammalian species, local estrogen dioactivity in duplicate 200-,ul samples of each incubate was biosynthesis is believed to be critically important for andro- separated by Sephadex LH-20 gel filtration (15). Protein gen-induced sexual differentiation of the brain (1, 13). The content of the cytosols was measured by the method of Bradford (16). The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" Abbreviations: E1, estrone; E2, estradiol. in accordance with 18 U.S.C. §1734 solely to indicate this fact. fTo whom reprint requests should be addressed. 513 Downloaded by guest on September 27, 2021 514 Neurobiology: MacLusky et al. Proc. Natl. Acad Sci. USA 83 (1986) The specificity of the binding reactions was studied by incubating the cytosol fractions with a single concentration of [3H]moxestrol (2nM) in the presence of unlabeled estradiol, 0. progesterone, cortisol, testosterone, or the synthetic estro- gen diethylstilbestrol-all added to a final concentration of 100 nM. 00 Aromatase complex activity was measured by a procedure mc based on the product isolation methods as described by Callard et al. (17) and Milewich et al. (18) with 4- 0.1 [3H]androstene-3,17-dione as substrate. Tissues were ho- o . mogenized with a motor-driven Teflon-glass homogenizer in 4 8 12 200 400 buffer (1 ml per 100 mg of tissue) containing 250 mM sucrose Bound, fmol/mg of protein and 50 mM sodium phosphate (pH 7.2). Aliquots ofthe tissue homogenates (1.8 ml; in duplicate in the case of the larger FIG. 1. High-affinity binding for 13H]moxestrol in cytosol frac- cerebral cortex samples) were incubated for 2 hr at 370C with tions prepared from the tissues ofa 6-day-old female rhesus monkey. 0.2 ml ofphosphate buffer containing an NADPH-generating The graph depicts Scatchard (21) plots of [3H]moxestrol binding in system (4 mM ATP/4 mM NADP/10 mM glucose 6- cytosol fractions prepared from the pooled hypothalamus and phosphate/4 units of dehydrogenase) preoptic area (A) and orbital cortex (A) (Left) and from the uterus (o) glucose-6-phosphate and pituitary (e) (Right). Equilibrium dissociation constants [Kd and 50 pmol of [1,2,6,7- H(N)]andros-4-ene-3,17-dione (New (nM)] and saturation binding capacities [NS (fmol/mg of cytosol England Nuclear; specific activity, 87 Ci/mmol). Ten micro- protein)] for all of the tissues examined were as follows: uterus, Kd grams each of unlabeled estradiol (E2) and estrone (E1) were = 1.14, NS = 392; pituitary, Kd = 0.66, NS = 97; hypothalamus- added to the incubates as an estrogen "trap" (13). At the end preoptic area, Kd = 0.41, NS = 12.4; hippocampus, Kd = 0.47, NS ofthe incubation, the reaction was stopped by adding 5 ml of = 3.0; orbital-prefrontal cortex, Kd = 0.48, NS = 3.15; anterior toluene containing 0.25% Triton X-100. Unlabeled "carrier" cingulate cortex, Kd = 0.62, NS = 4.2; dorsolateral-prefrontal steroids (200 ,ug each ofE1 and E2) and 14C-labeled E1 and E2 cortex, Kd = 0.48, NS = 3.2; somatosensory cortex, Kd = 0.61, ND (2000 dpm each) were added. The incubates were extracted = 4.1; motor cortex, Kd = 0.71, NS = 3.5. with 3 x 5 ml oftoluene. The combined toluene extracts were evaporated to dryness and subjected to phenolic partition to motor cortices from the postnatal animals, sufficient [3H]E1 separate the androgen and estrogen metabolites present. The was obtained for analysis by reverse isotopic dilution and estrogens were purified by TLC on silica gel GF254 plates crystallization to constant specific activity. The 3H:14C ratios (Brinkmann) in the system described by Ruh (19). The in the purified E1 fractions changed by less than 5% over separated E1 and E2 fractions were eluted from the TLC three successive crystallizations, confirming the identity of plates and then acetylated by overnight incubation at room the major estrogen product. The results of the aromatase temperature with 0.2 ml of pyridine and 0.2 ml of acetic complex measurements are summarized in Fig. 3. As expect- anhydride. The acetate fractions were chromatographed ed, the highest levels of estrogen formation were observed in once more on silica gel plates in methylene chloride/ethyl the hypothalamus-preoptic area. However, the amygdala, acetate, 98:2 (vol/vol). Radiochemical purity of the final the hippocampus, and several of the cortex samples also separated estrogen acetates was confirmed, wherever possi- contained measurable aromatase complex activity. Among ble, by crystallization to constant specific activity from the cortical samples, the highest levels ofaromatase complex acetone and hexane. Corrections for procedural losses were activity were found in the dorsolateral-prefrontal, orbital- made on the basis of the recoveries ofthe 14C standards. The prefrontal, anterior cingulate, and parietal cortices. The protein content of the homogenates was measured by the lowest levels were found in the somatosensory visual and method of Lowry et al. (20). motor cortices of the postnatal animals. Overall estrogen

RESULTS ._ 4-0 Cytosol [3H]moxestrol binding was studied only in the tissues 0. from the infant females. The results from the first animal are 15 summarized in Fig. 1. High-affinity saturable [3H]moxestrol 0 - binding sites were detected in all brain regions examined. The 0 12 apparent equilibrium dissociation constants for reaction of these sites with [3H]moxestrol were similar to those for uterine and pituitary cytosol estrogen receptors (0.3-1.1 nM). 4- Within the brain, highest levels of binding were observed in the hypothalamus-preoptic area, with fairly even lower x m

concentration throughout the cortical structures. The spec- l6 ificity of the binding reactions was examined in cytosols _ 3 prepared from the same regions of the second postnatal 3 female's brain. In all cases, binding of [3H]moxestrol ap- m peared to be estrogen-specific: thus, [3H]moxestrol binding F lTI was suppressed by coincubation with unlabeled E2 or N E P C T D N E P C T D diethylstilbestrol but was unaffected by progesterone, cortisol, or testosterone (Fig. 2). FIG. 2. Specificity of [3H]moxestrol binding in cytosols from Aromatase complex activity was detected in the majority hypothalamus-preoptic area (Left) and orbital-prefrontal cortex or (Right) from a 6-day-old female rhesus monkey. Cytosols were of the tissue specimens. There were no marked regional incubated with 2 nM [3H]moxestrol in the presence and absence of age-related differences in the relative proportions of E1 and the unlabeled competitors (all added to a final concentration of 100 E2 formed: in all tissues, the ratio of [3H]E1 to [3H]E2 was nM). Bars: E, estradiol; P, progesterone; C, cortisol; T, testosterone; approximately 9:1. In all brain regions from the prenatal D, diethylstilbestrol; N, control incubates, with no unlabeled com- female and in all brain regions except the somatosensory and petitor. Results are means ± SEM of triplicate determinations. Downloaded by guest on September 27, 2021 Neurobiology: MacLusky et al. Proc. Natl. Acad. Sci. USA 83 (1986) 515 (23) have proposed that aromatase complex activity may be * present transiently in some regions of the mammalian brain 0. during early development. The observations of Flores et al. 0 1oo0 (24) indicating that estrogen is not detectable in the cerebral w * cortex of the adult rhesus monkey after perfusion with 00 4-[3H]androstene-13,17-dione, despite the presence of sub- E stantial quantities of radiolabeled estrogen in the hypothal- A amus and limbic system, are also consistent with the hypoth- ~04)E A - 10 0 esis that aromatase complex activity in the primate cerebral cortex may decline life. A0 ~ ~ * during postnatal 4) Although local estrogen biosynthesis is believed to be an essential step in the mechanism of androgen-induced sexual 00 differentiation of the brain in rodents such as rats and mice 0 4- (1), the role of estrogen in the developing primate brain remains somewhat uncertain. In rhesus monkeys, although differentiation of masculine patterns of juvenile play and HPO AM HI DL OR AC PA V S M sexual behavior is clearly dependent on early sex differences in circulating testosterone levels (3), conversion of FIG. 3. Aromatase complex activity in the brain of the neonatal testosterone to estrogen is probably not required for this rhesus monkey. A, Results from a 2-day-old male; 0, results from a process to occur. Thus, prenatal administration of 5a- 5- and a 6-day-old female; *, results from a female fetus (153 days after conception). Shaded symbols for the male indicate that the dihydrotestosterone, which is not a substrate for the tissue samples from the left and right hemispheres were assayed aromatase complex, also appears to masculinize juvenile independently (shading of the left halfofthe symbol indicating results behavior (4). However, these findings do not preclude the from the left hemispheric sample, and vice versa). Histogram bars possiblity that normal differentiation may occur as a result of represent the mean values for each region, combining the data from the concerted effects of androgen and estrogen rather than all four animals. The means of the results from the two hemispheres either alone; nor do they rule out the possibility that estrogen in the 2-day-old male were used in calculating the overall mean may influence functions other than these particular behavior estrogen-formation rates. Results are expressed as fmol of estrogen patterns. Fox has suggested that the ratio of androgens to (E1 + E2) formed per mg of protein and are plotted semilogarithmi- estrogens may be important in the effects of cally. Bars: HPO, pooled hypothalamus and preoptic area, AM, determining amygdala; HI, hippocampus; DLM, cortical regions: dorsolate- these steroids on the developing central nervous system (25). ral-prefrontal (DL), orbital-prefrontal (OR), anterior cingulate (AC), Local estrogen formation could serve to augment the effects parietal (PA), visual (V), somatosensory (S), and motor (M). of circulating estrogens in specific target areas: although circulating estrogen levels are relatively high in primates during pregnancy, the fetus rapidly metabolizes estrogens to formation per mg of protein was considerably higher in the polar conjugates (26). This may effectively limit exposure of hypothalamus-preoptic area, amygdala, hippocampus, the fetal tissues to the circulating hormone in the same way orbital-prefrontal, somatosensory, and motor cortex sam- as binding to plasma proteins is believed to protect the tissues ples from the fetal female than in the homologous regions of developing rodents from maternal estrogens (1, 27). from the postnatal monkeys. Results in the latter two tissues The effects of locally synthesized estrogen may be partic- were particularly striking, in that estrogen formation was ularly important in primates in terms of the development of virtually undetectable in all of the postnatal somatosensory sex differences in cortical function. Several lines ofevidence and motor cortex specimens but was clearly measurable in suggest that estrogen may have maturational effect on the the fetal tissues. In the other brain regions, similar results cortex. In rodents, estrogen administration during the first were obtained with fetal and postnatal tissue samples. postnatal week accelerates cortical myelination (28) and causes precocious maturation of cerebral amino acid levels (29) and electroshock seizure thresholds (30). In the human DISCUSSION there is evidence that estrogen exposure during early life influences cerebral lateralization: young women exposed to These results shed light on the actions of gonadal steroids on diethylstilbestrol in utero show greater verbal response the developing primate brain. The [3H]moxestrol binding lateralization than their unexposed siblings (31), whereas data suggest that cytosol estrogen receptors are present at women with ovarian dysgenesis, who are deficient in ovarian high concentrations in the hypothalamus-preoptic area and hormones, show an even lower degree of lateralization than somewhat lower concentrations in several regions of the normal females (32). One previous report has indicated that cerebral cortex at the time of birth. This is consistent with a frontal cortex tissue from first-trimester human fetuses may recent report of putative cytosol estrogen receptors in the have some capacity for estrogen biosynthesis (33). In the hypothalamus-preoptic area and cerebral cortex of fetal present study, it is noteworthy that the highest levels of rhesus monkeys between 135 and 162 days of gestation (22). aromatization were observed in the orbital-prefrontal region The aromatization results further suggest hypothalamus-pre- of the cortex. Indeed, the levels of aromatase tended to be optic area, amygdala, hippocampus, and several regions of higher in all four areas of association cortex studied. These the cortex have the four areas are anatomatically interconnected and form a capacity to synthesize estrogens from neural system that may be important for cognitive processes circulating androgens. The disparity between the regional (34-36). Previous studies have shown that the behavioral distribution of aromatase complex activity in the effects of lesions of the orbital-prefrontal cortex in infant somatosensory and motor cortices of the prenatal and post- rhesus monkeys are sex-dependent, suggesting that this part natal animals further suggests that changes may occur in the of the brain develops more rapidly in males than in females levels and distribution of this enzyme during the perinatal (5). Administration of testosterone propionate to infant period. Whether this is in fact the case will have to be female monkeys masculinizes their response to orbital- determined by more extensive comparisons of monkeys at prefrontal lesions, indicating that the normal sex difference in different stages of gestation and during early postnatal life. this response is hormone-dependent and is sensitive to However, in this context it is noteworthy that Callard et al. gonadal steroids into postnatal life (6). The present observa- Downloaded by guest on September 27, 2021 516 Neurobiology: MacLusky et al. Proc. Natl. Acad Sci. USA 83 (1986) tions raise the possibility that these effects may at least in part 16. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254. be mediated through the actions of estrogen, synthesized 17. Callard, G. V., Petro, Z. & Ryan, K. J. (1978) Endocrinology locally from circulating androgens. 103, 2283-2290. 18. Milewich, L., George, F. W. & Wilson, J. D. (1977) Endocri- We are indebted to Clifford Hulburt for excellent technical nology 100, 187-196. assistance. This work was supported by a grant from the National 19. Ruh, T. S. (1976) J. Chromatogr. 121, 82-84. Institutes of Health (NS19610). 20. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 1. MacLusky, N. J. & Naftolin, F. (1981) Science 211, 1294- 21. Scatchard, G. (1949) Ann. N. Y. Acad. Sci. 51, 660-672. 1303. 22. Pomerantz, S. M., Fox, T. O., Sholl, S. A., Vito, C. C. & 2. Diamond, M. C., Dowling, G. A. & Johnson, R. E. (1980) Goy, R. W. (1985) Endocrinology 116, 83-89. Exp. Neurol. 71, 261-268. 23. Callard, G. V., Petro, Z. & Ryan, K. J. (1978) Am. Zool. 18, 3. Goy, R. W. & Resko, J. A. (1972) Recent Prog. Horm. Res. 511-523. 28, 707-733. 24. Flores, F., Naftolin, F., Ryan, K. J. & White, R. J. (1973) 4. Goy, R. W. (1978) in Recent Advances in Primatology, eds. Science 180, 1074-1075. Chivers, D. J. & Herbert, J. (Academic, New York), pp. 25. Fox, T. 0. (1975) Proc. Natl. Acad. Sci. USA 72, 4303-4307. 449-462. 26. Davidson, I. A., Stott, J. E., Longstaff, A., Abramovich, 5. Goldman, P. S., Crawford, H. T., Stokes, L. P., Galkin, D. R. & Pearson, C. K. (1983) J. Steroid Biochem. 18, T. W. & Rosvold, H. E. (1974) Science 186, 540-542. 525-529. 6. Goldman, P. S. & Brown, R. M. (1975) Neurosci. Abstr. 1, 5. 27. Diczfalusy, E., Cassner, O., Alonso, C. & de Miquel, M. 7. Critchley, M. (1970) The Dyslexic Child (Thomas, Springfield, (1961) Recent Prog. Horm. Res. 17, 147-206. IL), pp. 90-92. 28. Curry, J. J. & Heim, L. M. (1966) Nature (London) 209, 8. Maccoby, E. E. & Jacklin, C. N. (1974) The Psychology ofSex 915-916. Differences (Stanford Univ. Press, Stanford, CA). 29. Hudson, D. B., Vernadakis, A. & Timiras, P. S. (1970) Brain 9. McGlone, J. (1980) Behav. Brain Sci. 3, 215-263. Res. 23, 213-222. 10. Ehrhardt, A. A. & Meyer-Bahlburg, H. F. L. (1981) Science 30. Heim, L. M. & Timiras, P. S. (1963) Endocrinology 72, 211, 1312-1318. 598-606. 11. Hier, D. B. & Crowley, W. F., Jr. (1982) N. Engl. J. Med. 306, 31. Hines, M. (1982) Psychol. Bull. 92, 56-80. 1202-1227. 32. Gordon, H. W. & Galatzer, A. (1980) Psychoneuroendocrinol- 12. Geschwind, N. & Behan, P. (1982) Proc. Natl. Acad. Sci. USA ogy (Oxford) 5, 235-244. 79, 5097-5100. 33. Ryan, K. J., Naftolin, F., Reddy, V., Flores, F. & Petro, Z. 13. Naftolin, F., Ryan, K. J., Davies, I. J., Reddy, V. V., Flores, (1972) Am. J. Obstet. Gynecol. 114, 454-460. F., Petro, Z., Kuhn, M., White, R. J., Takaoka, Y. & Wolin, 34. Goldman-Rakic, P. S. & Schwartz, M. L. (1982) Science 216, L. (1975) Recent Prog. Horm. Res. 31, 295-319. 755-757. 14. Brown, R. M., Crane, A. M. & Goldman, P. S. (1979) Brain 35. Pandya, D. N., Dye, P. & Butters, N. (1971) Brain Res. 31, Res. 168, 133-150. 35-46. 15. MacLusky, N. J., Lieberburg, I. & McEwen, B. S. (1979) 36. Schwartz, M. L. & Goldman-Rakic, P. S. (1984) J. Comp. Brain Res. 178, 129-142. Neurol. 226, 403-420. Downloaded by guest on September 27, 2021